1. Field of the Invention
This invention relates to the field of cloning and expression of sialyltransferase enzymes. In particular, the preferred sialyltransferases are bacterial transferases obtained from, for example, Campylobacter jejuni. 
2. Background
Carbohydrates are now recognized as being of major importance in many cell-cell recognition events, notably the adhesion of bacteria and viruses to mammalian cells in pathogenesis and leukocyte-endothelial cell interaction through selectins in inflammation (Varki (1993) Glycobiology 3: 97-130). Moreover, sialylated glycoconjugates that are found in bacteria (Preston et al. (1996) Crit. Rev. Microbiol. 22:139-180; Reuter et al. (1996) Biol. Chem. Hoppe-Seyler 377:325-342) are thought to mimic oligosaccharides found in mammalian glycolipids to evade the host immune response (Moran et al. (1996) FEMS Immunol. Med. Microbiol. 16:105-115). Molecular mimicry of host structures by the saccharide portion of lipopolysaccharide (LPS) is considered to be a virulence factor of various mucosal pathogens, which use this strategy to evade a host immune response (Moran et al. (1996) FEMS Immunol. Med. Microbiol. 16: 105-115; Moran et al. (1996) J. Endotoxin Res. 3: 521-531).
One such pathogen, Campylobacter jejuni, is an important cause of acute gastroenteritis in humans (Skirrow (1977) Brit. Med. J. 2: 9-11). Epidemiological studies have shown that Campylobacter infections are more common in developed countries than Salmonella infections, and they are also an important cause of diarrheal diseases in developing countries (Ketley (1997) Microbiol. 143: 5-21). Moreover, C. jejuni infection has been implicated as a frequent antecedent to the development of Guillain-Barré syndrome, a form of neuropathy that is the most common cause of generalized paralysis (Ropper (1992) N. Engl. J. Med. 326: 1130-1136). The C. jejuni serotype most commonly associated with Guillian-Barré syndrome is O:19 (Kuroki et al. (1993) Ann. Neurol. 33: 243-247). The core oligosaccharides of low molecular weight LPS of O:19 strains exhibit molecular mimicry of several gangliosides (Aspinall et al. (1994) Biochemistry 33: 241-249; Aspinall et al. (1994) Biochemistry 33: 250-255). Terminal oligosaccharide moieties identical to those of GD1a, GD3, GM1 and GT1a gangliosides have been found in various O:19 strains. The significance of molecular mimicry as a virulence factor makes the identification of the genes involved in LPS synthesis and the study of their regulation of considerable interest for a better understanding of the pathogenesis mechanisms used by these bacteria.
The oligosaccharide structures involved in these and other processes are potential therapeutic agents, but they are time consuming and expensive to make by traditional chemical means. A very promising route to production of specific oligosaccharide structures is through the use of the enzymes which make them in vivo, the glycosyltransferases. Such enzymes can be used as regio- and stereoselective catalysts for the in vitro synthesis of oligosaccharides (Ichikawa et al. (1992) Anal. Biochem. 202: 215-238). Sialyltransferases are a group of glycosyltransferases that transfer sialic acid from an activated sugar nucleotide to acceptor oligosaccharides found on glycoproteins, glycolipids or polysaccharides. The large number of sialylated oligosaccharide structures has led to the characterization of many different sialyltransferases involved in the synthesis of various structures. Based on the linkage and acceptor specificity of the sialyltransferases studied so far, it has been determined that at least 13 distinct sialyltransferase genes are present in mammals (Tsuji et al. (1996) Glycobiology 6:v-vii).
Large scale enzymatic synthesis of oligosaccharides depends on the availability of sufficient quantities of the required glycosyltransferases. However, production of glycosyltransferases in sufficient quantities for use in preparing oligosaccharide structures has been problematic. Expression of many mammalian glycosyltransferases has been achieved involving expression in eukaryotic hosts which can involve expensive tissue culture media and only moderate yields of protein (Kleene et al. (1994) Biochem. Biophys. Res. Commun. 201: 160-167; Williams et al. (1995) Glycoconjugate J. 12: 755-761). Expression in E. coli has been achieved for mammalian glycosyltransferases, but these attempts have produced mainly insoluble forms of the enzyme from which it has been difficult to recover active enzyme in large amounts (Aoki et al. (1990) EMBO. J. 9:3171-3178; Nishiu et al. (1995) Biosci. Biotech. Biochem. 59 (9): 1750-1752). Furthermore, because of the biological activity of their products, mammalian sialyltransferases generally act in specific tissues, cell compartments and/or developmental stages to create precise sialyloglycans.
Bacterial sialyltransferases are not subject to the same constraints and can use a wider range of acceptors than that of the mammalian sialyltransferases. For instance, the α-2,6-sialyltransferase from Photobacterium damsela has been shown to transfer sialic acid to terminal galactose residues which are fucosylated or sialylated at the 2 or 3 position, respectively (Kajihara et al. (1996) J. Org. Chem. 61:8632-8635). Such an acceptor specificity has not been reported so far for mammalian sialyltransferases. Despite their importance as proven or potential virulence factors, as well as their potential use in synthesizing sialylated oligosaccharides of interest, few bacterial sialyltransferases have been cloned (Weisgerber et al. (1991) Glycobiol. 1:357-365; Frosch et al. (1991) Mol. Microbiol. 5:1251-1263; Gilbert et al. (1996) J. Biol. Chem. 271:28271-28276) or purified (Yamamoto et al. (1996) J. Biochem. 120:104-110). The α-2,8-sialyltransferases involved in the synthesis of the polysialic acid capsules have been cloned and expressed from both Escherichia coli (Weisgerber et al. (1991) Glycobiol. 1:357-365) and N. meningitidis (Frosch et al. (1991) Mol. Microbiol. 5:1251-1263). Glycosyltransferases from N. gonorrhoeae which are involved in the synthesis of lipooligosaccharide (LOS) have been cloned (U.S. Pat. No. 5,545,553).
Thus, bacterial sialyltransferases would be useful in a number of applications, such as the synthesis of desired oligosaccharides with biological activity. Identification and characterization of new bacterial sialyltransferases would thus be useful in the development of these technologies. The present invention fulfills this and other needs.